Microsoft Certified Azure AI Engineer Associate (1 科 Microsoft 雲端人工智能) 國際認可證書課程

人工智能是什麼？
人工智能 (AI) 是展現一種或多種接近人類能力的軟體：從視覺感知到分析學習、解決問題和作出決策等等。

人工智能 (AI) 在我們日常使用的軟體應用程式中越來越普遍；包括我們家裡和手機中的數位助理、車輛中的自動駕駛技術，以及幫助我們完成工作的生產力應用程式。

透過 AI，機器可以分析資料、文字和影像、理解語音、以自然方式互動，以及使用資料進行預測，例如電腦何時需要維護，或客戶可能想要購買何種類型的產品。

Azure 是彈性的平台，可讓您的組織充份利用 Microsoft AI 及 Open AI 服務。當您使用以 Azure 雲端為運行基礎的新一代智慧型應用程式時，您仍然可以自行決定放置企業資料的位置，例如雲端 (cloud)、內部部署 (on-premises)，以及智慧邊緣 (intelligent edge)。

Azure 能讓您擁有：

• 視覺感知 - 使用電腦視覺功能來處理影像、視訊串流和即時攝影機輸入影片。
• 文字分析與對話 - 使用自然語言處理 (Natural Language Processing & Large Language Models) 不僅可以閱讀，還可以產生真實的回應並從文件中提取要點及作出總結的能力。
• 語音 - 辨識語音輸入並合成語音輸出的能力。語音功能與自然語言處理分析的能力相結合，實現了一種被稱為對話式 AI 的人機交談形式。
• 作出決策 - 利用過去的經驗和學習 (Machine Learning) 來評估情況並採取適當行動的能力。例如，識別感應器不尋常讀數 (Multivariate Multiple Regression) 並採取自動化回應操作來防止故障或系統損壞。
• Microsoft 及 Open AI 的人工智能技術。
• 最完整的合規性和安全性產品。

AI 應用程式是融合了 AI 功能的 Web 或行動應用程式，例如視覺辨識或語言處理。具體例子如使用臉部辨識作為行動裝置登入程序一部份的金融服務應用程式，就是融合了 AI 功能的行動應用程式。

AI 代理程式則是使用 AI 功能與真人使用者互動的電腦程式。例如，某公司使用聊天 Bot 管理即時的 Web 式客戶服務要求，即為 AI 代理程式的範例。

透過租用或購買 Microsoft 的智慧型應用程式和代理程式的 AI 模型 (Pre-trained models)，令您可以親自制定您的商業應用程式來提升您的產品及服務水平！

Microsoft 利用 Azure 認知服務讓您的應用程式有突破性體驗。各種產業以及各種不同的產品，例如 Microsoft 365、Xbox 與 Bing，現在都使用領先的 AI 模型。

關於 AI Engineer 認證
Azure AI Engineer Associate 認證的應試者應具有機器學習 (ML) 和人工智慧 (AI) 概念，以及相關 Microsoft Azure 服務的實務知識。這認證是一個展示常見 AI 工作負載以及如何在 Azure 上實際運作 AI 技巧的機會。

Microsoft Certified: Azure AI Engineer Associate

About the course
通過本課程，您將學習管理、部署和利用 Azure AI 認知服務 (Cognitive services)、AI 搜尋 (Search and Indexing)，並了解如何使用基於 REST 的 API 在 Azure 上建立電腦視覺、語言分析、知識挖掘、智慧搜尋和 Open AI 的生成式解決方案。

我們的資深講師 Larry Chan 將為您提供各種 Microsoft Azure 認知服務 (Cognitive services) 及 Open AI 相關產品的建議和運用技巧。

課程名稱：	Microsoft Certified Azure AI Engineer Associate (1 科 Microsoft 雲端人工智能) 國際認可證書課程 - 簡稱：Azure AI Engineer Training Course
課程時數：	42 小時 (共 14 堂，共 1 科)
適合人士：	有志考取 Microsoft Certified: Azure AI Engineer Associate 證書人士
授課語言：	以廣東話為主，輔以英語
課程筆記：	本中心導師親自編寫英文為主筆記，而部份英文字附有中文對照。

1. 模擬考試題目：	本中心為學員提供模擬考試題目，每條考試題目均附有標準答案。
2. 時數適中：	本中心的 Microsoft Certified Azure AI Engineer Associate (1 科 Microsoft 雲端人工智能) 國際認可證書課程時數適中，有 42 小時。 令學員能真正了解及掌握課程內容，而又能於 3 個月內考獲以下 1 張國際認可證書: Microsoft Certified Azure AI Engineer Associate
3. 導師親自編寫筆記：	由本中心已擁有五項 MCITP ， 十多項 MCTS，MCSA 及 MCSE 資格，並有教授 Microsoft 相關課程 24年以上經驗的資深導師 Larry Chan 親自編寫筆記，絕對適合考試及實際管理之用，令你無須「死鋤」如字典般厚及不適合香港讀書格調的書本。
4. 一人一機上課：	本課程以一人一機模式上課。
5. 免費重讀：	傳統課堂學員可於課程結束後三個月內免費重看課堂錄影。

Microsoft 已公佈考生只要通過以下 1 個 Azure AI 相關科目的考試，便可獲發 Microsoft Certified Azure AI Engineer Associate 國際認可證書:

考試編號	科目名稱
AI-102	Designing and Implementing a Microsoft Azure AI Solution

本中心為Microsoft指定的考試試場。報考時請致電本中心，登記欲報考之科目考試編號、考試日期及時間 (最快可即日報考)。臨考試前要出示身份證及繳付每科HK$943之考試費。 考試題目由澳洲考試中心傳送到你要應考的電腦，考試時以電腦作答。所有考試題目均為英文，而大多數的考試題目為單項及多項選擇題，其餘則為實戰題。 考試合格後會收到來自Microsoft的作實電郵，並進入該電郵內的連結，登入 Microsoft Credentials Dashboard 下載您的證書。 考試不合格便可重新報考，不限次數。欲知道作答時間、題目總數、合格分數等詳細考試資料，可瀏覽本中心網頁 "各科考試分數資料"。

課程名稱：Microsoft Certified Azure AI Engineer Associate (1 科 Microsoft 雲端人工智能) 國際認可證書課程 - 簡稱：Azure AI Engineer Training Course

AI-102 Designing and Implementing a Microsoft Azure AI Solution

1. Prepare to develop AI solutions on Azure
1.1 About Artificial Intelligence
1.2 AI-related terms
1.2.1 Data science
1.2.2 Machine learning
1.2.3 Artificial intelligence
1.3 Considerations for AI Engineers
1.3.1 Model training and inferencing
1.3.2 Probability and confidence scores
1.3.3 Responsible AI and ethics
1.4 Considerations for responsible AI
1.4.1 Fairness
1.4.2 Reliability and safety
1.4.3 Privacy and security
1.4.4 Inclusiveness
1.4.5 Transparency
1.4.6 Accountability
1.5 Capabilities of Azure Machine Learning
1.6 Regression with Azure Machine Learning
1.6.1 Create an Azure Machine Learning workspace
1.6.2 Create compute resources
1.6.3 Explore data
1.6.4 Train a machine learning model
1.6.5 Review the best model
1.6.6 Deploy a model as a service
1.6.7 Test the deployed service
1.7 Capabilities of Azure AI Services
1.8 capabilities of the Azure OpenAI Service
1.9 Capabilities of Azure AI Search

2. More about Consuming Azure AI Services
2.1 Various Azure AI Services
2.2 Provision an Azure AI services resource
2.3 Identify endpoints and keys
2.4 Use a REST API
2.4.1 Example of using Python to construct and send a request
2.4.2 The Roles of HTTP, APIs, and REST
2.4.3 The Request
2.4.4 The Response
2.4.5 How to Use Python Requests with REST APIs
2.4.6 How to Access REST Headers
2.4.7 How to Authenticate to a REST API
2.4.8 How to Handle HTTP Errors With Python Requests
2.4.9 How to Check for HTTP Errors With Python Requests
2.4.10 TooManyRedirects
2.4.11 ConnectionError
2.4.12 Timeout
2.5 Use an SDK
2.5.1 Microsoft Azure SDK for Python
2.6 Lab Exercise of Provisioning Azure AI Services
2.6.1 Get Started with Azure AI Services
2.6.2 Clone the repository in Visual Studio Code
2.6.3 Provision an Azure AI Services resource
2.6.4 Use a REST Interface
2.6.5 Using an SDK

3. Securing Azure AI services
3.1 Consider authentication
3.1.1 Regenerate keys
3.1.2 Protect keys with Azure Key Vault
3.1.3 Token-based authentication
3.1.4 Microsoft Entra ID authentication
3.1.5 Authenticate using managed identities
3.2 Implement network security for Azure AI Services
3.2.1 Apply network access restrictions
3.3 Manage Azure AI Services Security
3.3.1 Clone the repository in Visual Studio Code
3.3.2 Provision an Azure AI Services resource if required
3.3.3 Manage authentication keys
3.3.4 Secure key access with Azure Key Vault
3.3.5 Create a key vault and add a secret
3.3.6 Create a service principal
3.3.7 Use the service principal in an application

4. Monitor Azure AI services
4.1 Monitor cost
4.1.1 Plan costs for AI services
4.1.2 View costs for AI services
4.1.3 Costs might accrue before resource deletion
4.1.4 Costs might accrue after resource deletion
4.1.5 Create budgets
4.2 Creating Alerts for Azure AI Services
4.2.1 Alert rules
4.3 Viewing metrics
4.3.1 View metrics in the Azure portal
4.3.2 Add metrics to a dashboard
4.4 Monitoring Azure AI Services and configuring Alerts
4.4.1 Monitor Azure AI Services
4.4.2 Clone the repository in Visual Studio Code
4.4.3 Provision an Azure AI Services resource
4.4.4 Configure an alert
4.4.5 Visualize a metric
4.5 Manage diagnostic logging
4.5.1 Create resources for diagnostic log storage
4.5.2 Configure diagnostic settings
4.5.3 View log data in Azure Log Analytics

5. Deploy Azure AI services in containers
5.1 Understanding Azure Containers
5.1.1 What is a container?
5.1.2 Container deployment
5.2 An example Azure Container Instance configuration
5.2.1 Creating Azure Container Instance
5.2.2 Viewing Container Logs
5.3 More about using Azure AI Services containers
5.3.1 Azure AI services container images
5.3.2 Azure AI services container configuration
5.4 Deploying and Configuring Azure AI Services Container
5.4.1 Clone the repository in Visual Studio Code
5.4.2 Provision an Azure AI Services resource
5.4.3 Deploy and run a Text Analytics container
5.4.4 Using the Azure AI Services container
5.5 Benefits of Using Azure AI Services container

6. Analyze images
6.1 Provision an Azure AI Vision resource
6.2 Analyze an image
6.3 Generate a smart-cropped thumbnail and remove background
6.3.1 Remove image background
6.4 Analyze images with Azure AI Vision
6.4.1 Clone the repository for this course
6.4.2 Provision an Azure AI Services resource
6.4.3 Prepare to use the Azure AI Vision SDK
6.4.4 View the images you will analyze
6.4.5 Analyze an image to suggest a caption
6.4.6 Get suggested tags for an image
6.4.7 Detect and locate objects in an image
6.4.8 Detect and locate people in an image
6.4.9 Remove the background or generate a foreground matte of an image

7. Image classification with Custom Azure AI Vision models
7.1 Understand custom model types
7.1.1 Image classification
7.1.2 Object detection
7.1.3 Product recognition
7.2 Create a custom project
7.2.1 Components of a custom Vision project
7.2.2 COCO files
7.2.3 Creating your dataset
7.3 Classify images with an Azure AI Vision custom model
7.3.1 Clone the repository for this chatper
7.3.2 Provision Azure resources
7.3.3 Create a custom model training project
7.3.4 Test your custom model

8. Detect, Analyze, and Recognize faces
8.1 Identify options for face detection analysis and identification
8.1.1 The Azure AI Vision service
8.1.2 The Face service
8.2 Considerations for face analysis
8.3 Detect faces with the Azure AI Vision service
8.4 Capabilities of the face service
8.5 Compare and match detected faces
8.6 Implement facial recognition
8.7 Detect and Analyze Faces
8.7.1 Clone the repository for this course
8.7.2 Provision an Azure AI Services resource
8.7.3 Prepare to use the Azure AI Vision SDK
8.7.4 View the image you will analyze
8.7.5 Detect faces in an image
8.7.6 Prepare to use the Face SDK
8.7.7 Detect and analyze faces

9. Read Text in Images and Documents
9.1 Azure AI Vision options for Reading Text
9.1.1 Image Analysis Optical character recognition (OCR):
9.1.2 Document Intelligence:
9.2 Using the Read API
9.3 Reading Text in Images
9.3.1 Clone the repository for this course
9.3.2 Provision an Azure AI Services resource
9.3.3 Use the Azure AI Vision SDK to read handwritten text from an image

10. Analyze video
10.1 Azure Video Indexer capabilities
10.2 Extract custom insights
10.3 Using Video Analyzer widgets and APIs
10.3.1 Azure Video Indexer widgets
10.3.2 Azure Video Indexer API
10.3.3 Deploy with ARM template
10.4 Analyze video with Video Indexer
10.4.1 Clone the repository for this course
10.4.2 Upload a video to Video Indexer
10.4.3 Review video insights
10.4.4 Search for insights
10.4.5 Use Video Indexer widgets
10.4.6 Use the Video Indexer REST API

11. Analyze text with Azure AI Language
11.1 Provision an Azure AI Language resource
11.2 Detect language
11.3 Extract key phrases
11.4 Analyze sentiment
11.5 Extract entities
11.6 Extract linked entities
11.7 Analyze text with Azure AI Language service
11.7.1 Provision an Azure AI Language resource
11.7.2 Prepare to develop an app in Visual Studio Code
11.7.3 Configure your application
11.7.4 Add code to detect language
11.7.5 Add code to evaluate sentiment
11.7.6 Add code to identify key phrases
11.7.7 Add code to extract entities
11.7.8 Add code to extract linked entities

12. Question Answering Solutions
12.1 About Question Answering
12.2 Compare question answering to Azure AI Language understanding
12.3 Creating a Knowledge Base
12.4 Multi-Turn Conversation
12.5 Testing and publishing a Knowledge Base
12.5.1 Testing a knowledge base
12.5.2 Deploying a knowledge base
12.6 Consuming a Knowledge Base
12.7 Improve question answering performance
12.7.1 Use active learning
12.7.2 Create your question and answer pairs
12.7.3 Review suggestions
12.7.4 Define synonyms
12.8 A Lab exercise of Creating a Question Answer Solution
12.8.1 Provision an Azure AI Language resource
12.8.2 Create a question answering project
12.8.3 Add sources to the knowledge base
12.8.4 Edit the knowledge base
12.8.5 Train and test the knowledge base
12.8.6 Deploy the knowledge base
12.8.7 Prepare to develop an app in Visual Studio Code
12.8.8 Configure your application
12.8.9 Add code to the application

13. Conversational Language Understanding Model
13.1 Prebuilt Capabilities of the Azure AI Language service
13.1.1 Pre-configured features
13.1.2 Learned features
13.2 Azure Resources for building a Conversational Language Understanding model
13.2.1 Build your model
13.2.2 Use Language Studio
13.2.3 Use the REST API
13.2.4 Authentication
13.2.5 Request deployment
13.2.6 Get deployment status
13.2.7 Query your model
13.2.8 Query using the REST API
13.2.9 Sample response
13.3 Define intents, utterances, and entities
13.4 Use patterns to differentiate similar utterances
13.5 Employing Pre-built Entity Components
13.6 Train, test, publish, and review a conversational language understanding model
13.7 Building an Azure AI services conversational language understanding model
13.7.1 Create a language understanding model with the Language service
13.7.2 Provision an Azure AI Language resource
13.7.3 Create a conversational language understanding project
13.7.4 Create intents
13.7.5 Label each intent with sample utterances
13.7.6 Train and test the model
13.7.7 Add entities
13.7.8 Add a learned entity
13.7.9 Add a list entity
13.7.10 Add a prebuilt entity
13.7.11 Retrain the model
13.7.12 Use the model from a client app
13.7.13 Prepare to develop an app in Visual Studio Code
13.7.14 Configure your application
13.7.15 Add code to the application

14. Custom Text Classification Solution
14.1 Types of Classification Projects
14.1.1 Single vs. multiple label projects
14.1.2 Evaluating and improving your model
14.1.3 API payload
14.2 Concepts of Building a Text Classification Projects
14.2.1 Azure AI Language project life cycle
14.2.2 Splitting datasets for training
14.2.3 Deployment options
14.2.4 Using the REST API
14.2.5 Submit initial request
14.2.6 Get training status
14.2.7 Consuming a deployed model
14.2.8 Get classification results
14.3 Custom Text Classification Lab Exercise
14.3.1 Provision an Azure AI Language resource
14.3.2 Upload sample articles
14.3.3 Create a custom text classification project
14.3.4 Label your data
14.3.5 Train your model
14.3.6 Evaluate your model
14.3.7 Deploy your model
14.3.8 Prepare to develop an app in Visual Studio Code
14.3.9 Configure your application
14.3.10 Add code to classify documents
14.3.11 Test your application

15. Custom Named Entity Recognition
15.1 Concepts about Custom Named Entity Recognition (NER)
15.1.1 Custom vs built-in NER
15.1.2 Azure AI Language project life cycle
15.1.3 Considerations for data selection and refining entities
15.1.4 How to extract entities
15.1.5 Project limits
15.2 Label your data
15.2.1 How to label your data
15.3 Train and evaluate your model
15.3.1 How to interpret metrics
15.3.2 Confusion matrix
15.4 Extract Custom Entities Lab Exercise
15.4.1 Provision an Azure AI Language resource
15.4.2 Upload sample ads
15.4.3 Create a custom named entity recognition project
15.4.4 Label your data
15.4.5 Train your model
15.4.6 Evaluate your model
15.4.7 Deploy your model
15.4.8 Prepare to develop an app in Visual Studio Code
15.4.9 Configure your application
15.4.10 Add code to extract entities
15.4.11 Test your application

16. Translate text with Azure AI Translator service
16.1 Provision an Azure AI Translator resource
16.1.1 Azure resource for Azure AI Translator
16.2 Concepts on language detection, translation, and transliteration
16.2.1 Language detection
16.2.2 Translation
16.2.3 Transliteration
16.3 Specify translation options
16.3.1 Word alignment
16.3.2 Sentence length
16.3.3 Profanity filtering
16.4 Define custom translations
16.4.1 How to call the API
16.4.2 Response returned
16.5 Translate text with the Azure AI Translator service
16.5.1 Provision an Azure AI Translator resource
16.5.2 Prepare to develop an app in Visual Studio Code
16.5.3 Configure your application
16.5.4 Add code to translate text
16.5.5 Test your application

17. Create speech-enabled apps with Azure AI services
17.1 Provision an Azure resource for speech
17.2 Using the Azure AI Speech to Text API
17.2.1 Using the Azure AI Speech SDK
17.3 Using the text to speech API
17.3.1 Using the Azure AI Speech SDK
17.4 Configure audio format and voices
17.4.1 Audio format
17.4.2 Voices
17.5 Using Speech Synthesis Markup Language
17.6 Creating a Speech-enabled app
17.6.1 Recognize and synthesize speech
17.6.2 Provision an Azure AI Speech resource
17.6.3 Prepare to develop an app in Visual Studio Code
17.6.4 Configure your application
17.6.5 Add code to use the Azure AI Speech SDK
17.6.6 Add code to recognize speech
17.6.7 If you have a working microphone
17.6.8 Alternatively, use audio input from a file
17.6.9 Add code to process the transcribed command
17.6.10 Synthesize speech
17.6.11 Use a different voice
17.6.12 Use Speech Synthesis Markup Language

18. Translate speech with the Azure AI Speech service
18.1 Provision an Azure resource for speech translation
18.2 Translate speech to text
18.3 Synthesize translations
18.3.1 Event-based synthesis
18.3.2 Manual synthesis
18.4 Speed Translation Lab Exercise with Azure AI Speech service
18.4.1 Provision an Azure AI Speech resource
18.4.2 Prepare to develop an app in Visual Studio Code
18.4.3 Configure your application
18.4.4 Add code to use the Speech SDK
18.4.5 Implement speech translation
18.4.6 If you have a working microphone
18.4.7 Alternatively, use audio input from a file
18.4.8 Run the program
18.4.9 Synthesize the translation to speech

19. Azure AI Search solution
19.1 Manage Capacity
19.1.1 Service tiers and capacity management
19.1.2 Replicas and partitions
19.2 Search Components
19.2.1 Data source
19.2.2 Skillset
19.2.3 Indexer
19.2.4 Index
19.3 The indexing process
19.4 Search an index
19.4.1 Full text search
19.5 Filtering and Sorting
19.5.1 Filtering results
19.5.2 Filtering with facets
19.5.3 Sorting results
19.6 Enhance the index
19.6.1 Search-as-you-type
19.6.2 Custom scoring and result boosting
19.6.3 Synonyms
19.7 Creating an Azure AI Search Solution
19.7.1 Create Azure resources
19.7.2 Create an Azure AI Search resource
19.7.3 Create an Azure AI Services resource
19.7.4 Create a storage account
19.7.5 Prepare to develop an app in Visual Studio Code
19.7.6 Upload Documents to Azure Storage
19.7.7 Index the documents
19.7.8 Search the index
19.7.9 Explore and modify definitions of search components
19.7.10 Get the endpoint and key for your Azure AI Search resource
19.7.11 Review and modify the skillset
19.7.12 Review and modify the index
19.7.13 Review and modify the indexer
19.7.14 Use the REST API to update the search solution
19.7.15 Query the modified index
19.7.16 Create a search client application
19.7.17 Get the endpoint and keys for your search resource
19.7.18 Prepare to use the Azure AI Search SDK
19.7.19 Explore code to search an index
19.7.20 Explore code to render search results
19.7.21 Run the web app

20. Custom Skill for Azure AI Search
20.1 Create a custom skill
20.1.1 Input Schema
20.1.2 Output schema
20.2 Add a custom skill to a skillset
20.3 Implementing a Custom Skill for Azure AI Search lab exercise
20.3.1 Prepare to develop an app in Visual Studio Code
20.3.2 Create Azure resources
20.3.3 Create a search solution
20.3.4 Search the index
20.3.5 Create an Azure Function for a custom skill
20.3.6 Add the custom skill to the search solution
20.3.7 Search the index

21. Knowledge Store with Azure AI Search
21.1 Knowledge stores
21.2 Projections
21.2.1 Using the Shaper skill
21.3 Defining a Knowledge Store
21.4 Creating a Knowledge Store with Azure AI Search
21.4.1 Prepare to develop an app in Visual Studio Code
21.4.2 Create Azure resources
21.4.3 Create a search solution
21.4.4 Prepare JSON for REST operations
21.4.5 Submit REST requests
21.4.6 View the knowledge store
21.4.7 View object projections
21.4.8 View file projections
21.4.9 View table projections

22. Enrich your data with Azure AI Language
22.1 Available features of Azure AI Language
22.1.1 Azure AI Language features
22.1.2 Classify text
22.1.3 Understand questions and conversational language
22.1.4 Extract information
22.1.5 Summarize text
22.1.6 Translate text
22.1.7 Testing and using preconfigured language features
22.1.8 Create, train, and deploy a conversation language understanding model
22.2 Enriching a search index in Azure AI Search with custom classes and Azure AI Language
22.2.1 Store your data
22.2.2 Create your Azure AI Language project
22.2.3 Train your classification model
22.2.4 Create search index
22.2.5 Create an Azure function app
22.2.6 Update your Azure AI Search solution
22.2.7 Add a field to an existing index
22.2.8 Edit the custom skillset
22.2.9 Map the output from the function app into the index
22.3 Lab Exercise of Enriching an AI Search Index with Custom Classes
22.3.1 Set up your development environment with Python, VS Code and VS Code Extensions
22.3.2 Set up your Azure resources
22.3.3 Deploy a pre-built ARM template
22.3.4 Upload sample data to train language services
22.3.5 Create a language resource
22.3.6 Create a custom text classification project in Language Studio
22.3.7 Train your custom text classification AI model
22.3.8 Deploy your custom text classification AI model
22.3.9 Create an Azure AI Search index
22.3.10 Import documents into Azure AI Search
22.3.11 Create a function app to enrich your search index
22.3.12 Deploy your local function app to Azure
22.3.13 Test your remote function app
22.3.14 Add a field to your search index
22.3.15 Edit the custom skillset to call your function app
22.3.16 Edit the field mappings in the indexer
22.3.17 Test your enriched search index

23. Implement advanced search features in Azure AI Search
23.1 Improve the ranking of a document with term boosting
23.1.1 Search an index
23.1.2 Write a simple query
23.1.3 Enable the Lucene Query Parser
23.1.4 Boost search terms
23.2 Improve the relevance of results by adding scoring profiles
23.2.1 How search scores are calculated
23.2.2 Improve the score for more relevant documents
23.2.3 Add a weighted scoring profile
23.2.4 Use functions in a scoring profile
23.3 Improve an index with analyzers and tokenized terms
23.3.1 Analyzers in AI Search
23.3.2 What is a custom analyzer?
23.3.3 Tokenizers
23.3.4 Token filters
23.3.5 Create a custom analyzer
23.3.6 Test a custom analyzer
23.3.7 Use a custom analyzer for a field
23.4 Enhance an index to include multiple languages
23.4.1 Add language specific fields
23.4.2 Limit the fields for a language
23.4.3 Enrich an index with multiple languages using Azure AI Services
23.5 Improve search experience by ordering results by distance from a given reference point
23.5.1 What are geo-spatial functions?
23.5.2 Use the geo.distance function
23.5.3 Using the geo.intersects function
23.6 Implementing Enhancements to search results
23.6.1 Create Azure resources
23.6.2 Import sample data into the search service
23.6.3 Create an Azure AI Service to support translations
23.6.4 Add a translation enrichment
23.6.5 Change the field to store translated text
23.6.6 Update the skillset to translate the correct field in the document
23.6.7 Test the updated index
23.6.8 Add a scoring profile to improve search results
23.6.9 Test the updated index

24. Build an Azure Machine Learning custom skill for Azure AI Search
24.1 Concepts on Using a custom Azure Machine Learning skillset
24.1.1 Custom Azure Machine Learning skill schema
24.2 Enrich a search index using an Azure Machine Learning model
24.2.1 Create an AML workspace
24.2.2 Create and train a model in Azure Machine Learning studio
24.2.3 Alter how the model works to allow it to be called by the AML custom skill
24.2.4 Create an endpoint for your model to use
24.2.5 Connect the AML custom skill to the endpoint
24.3 Enriching a search index using Azure Machine Learning model
24.3.1 Create an Azure Machine Learning workspace
24.3.2 Create a regression training pipeline
24.3.3 Create an inference cluster for the endpoint
24.3.4 Register your trained model
24.3.5 Edit the scoring script to respond to Azure AI Search correctly
24.3.6 Create a custom environment
24.3.7 Deploy the model with the updated scoring code
24.3.8 Test your trained model’s endpoint
24.3.9 Integrate an Azure Machine Learning model with Azure AI Search
24.3.10 Create a test file
24.3.11 Create an Azure AI Search resource
24.3.12 Add cognitive skills
24.3.13 Add the AML Skill to the skillset
24.3.14 Update the output field mappings
24.3.15 Test index enrichment

25. Maintaining Azure AI Search solution
25.1 Manage security of an Azure AI Search solution
25.1.1 Overview of security approaches
25.1.2 Data encryption
25.1.3 Secure inbound traffic
25.1.4 Authenticate requests to your search solution
25.1.5 Secure outbound traffic
25.1.6 Secure data at the document-level
25.2 Optimize performance of an Azure AI Search solution
25.2.1 Measure your current search performance
25.2.2 Check if your search service is throttled
25.2.3 Check the performance of individual queries
25.2.4 Optimize your index size and schema
25.2.5 Improve the performance of your queries
25.2.6 Use the best service tier for your search needs
25.3 Manage costs of an Azure AI Search solution
25.3.1 Estimate your search solutions baseline costs
25.3.2 Understand the billing model
25.3.3 Tips to reduce the cost of your search solution
25.3.4 Manage search service costs using budgets and alerts
25.4 Improve reliability of an Azure AI Search solution
25.4.1 Make your search solution highly available
25.4.2 Distribute your search solution globally
25.4.3 Back up options for your search indexes
25.5 Monitor an Azure AI Search solution
25.5.1 Monitor Azure AI Search in Azure Monitor
25.5.2 Use metrics to see diagnostic data visually
25.5.3 Write Kusto queries against your search solutions logs
25.5.4 Create alerts to be notified about common search solution issues
25.6 Concept for Debugging Search issues
25.6.1 Explore how to use the Debug Session tool in Azure AI Search
25.6.2 Debug a skillset with Debug Sessions
25.6.3 Explore and edit a skill
25.6.4 Validate the field mappings
25.7 Debug search issues lab exercise
25.7.1 Create your search solution
25.7.2 Import sample data
25.7.3 Use a debug session to resolve warnings on your indexer
25.7.4 Resolve the warning on the indexer

26. Semantic Ranking in Azure AI Search
26.1 Semantic Ranking concepts
26.1.1 BM25 ranking function
26.1.2 Semantic ranking
26.1.3 Semantic captions and answers
26.1.4 How semantic ranking works
26.1.5 Semantic ranking advantages
26.1.6 Semantic ranking limitations
26.1.7 Semantic ranking pricing
26.2 Set up semantic ranking
26.2.1 Enable semantic ranking
26.2.2 Configure semantic ranking
26.3 Lab Exercise of Semantic Ranker
26.3.1 Enable semantic ranker
26.3.2 Import a sample index
26.3.3 Configure semantic ranking

27. Vector search and retrieval in Azure AI Search
27.1 Concepts about Vector Search
27.2 Prepare your search
27.2.1 Check your index has vector fields
27.2.2 Convert a query input into a vector
27.3 Understand embedding
27.3.1 Embedding models
27.3.2 Embedding space
27.4 Using the REST API to run vector search queries
27.4.1 Set up your project
27.4.2 Create an Index
27.4.3 Upload Documents
27.4.4 Run Queries

28. Planning an Azure AI Document Intelligence solution
28.1 Understanding AI Document Intelligence
28.1.1 Azure AI Document Intelligence
28.1.2 Responsible use of AI
28.1.3 Using models with Azure AI Document Intelligence
28.1.4 Azure AI Document Intelligence and Azure AI Vision
28.1.5 Azure AI Document Intelligence tools
28.2 Azure AI Document Intelligence resources
28.2.1 Create an Azure AI Document Intelligence resource
28.2.2 Connect to Azure AI Document Intelligence
28.3 Choose a model type
28.3.1 Prebuilt models
28.3.2 General document analysis models
28.3.3 Specific document type models
28.3.4 Custom models
28.3.5 Composed models

29. Using Prebuilt Document Intelligence models
29.1 Prebuilt Models
29.1.1 Features of prebuilt models
29.1.2 Input requirements
29.1.3 Try out prebuilt models with Azure AI Document Intelligence Studio
29.1.4 Calling prebuilt models by using APIs
29.2 Using the General Document, Read, and Layout models
29.2.1 Using the read model
29.2.2 Using the general document model
29.2.3 Using the layout model
29.3 Using Financial, ID, and Tax models
29.3.1 Using the invoice model
29.3.2 Using the receipt model
29.3.3 Using the ID document model
29.3.4 Using the business card model
29.3.5 Using other prebuilt models
29.4 Lab Exercise - Analyze a document using Azure AI Document Intelligence
29.4.1 Create an Azure AI Document Intelligence resource
29.4.2 Use the Read model
29.4.3 Prepare to develop an app in Visual Studio Code
29.4.4 Configure your application
29.4.5 Add code to use the Azure Document Intelligence service

30. Extracting data from forms with Azure Document intelligence
30.1 Role of Azure Document Intelligence
30.1.1 Azure Document Intelligence service components
30.1.2 Decide what component of Azure Document Intelligence to use
30.2 Training custom models
30.3 Use Azure Document Intelligence models
30.3.1 Using the API
30.3.2 Understanding confidence scores
30.4 Use the Azure Document Intelligence Studio
30.4.1 Build Document analysis model projects
30.4.2 Build prebuilt model projects
30.4.3 Build custom model projects
30.5 Lab Exercise of Extracting data from custom forms
30.5.1 Prepare to develop an app in Visual Studio Code
30.5.2 Create a Azure AI Document Intelligence resource
30.5.3 Gather documents for training
30.5.4 Train the model using Document Intelligence Studio
30.5.5 Test your custom Document Intelligence model

31. Composed Document intelligence model
31.1 Concepts of Composed Models
31.1.1 Using Composed models
31.1.2 Custom model compatibility
31.2 Assemble composed models
31.2.1 Create a composed model in Document Intelligence Studio
31.2.2 Create a composed model in code
31.3 Exercise of Creating a Composed Model
31.3.1 Run Cloud Shell
31.3.2 Set up resources
31.3.3 Create the 1040 Forms custom model
31.3.4 Label the 1040 Forms custom model
31.3.5 Train the 1040 Forms custom model
31.3.6 Create the 1099 Forms custom model
31.3.7 Label the 1099 Forms custom model
31.3.8 Train the 1099 Forms custom model
31.3.9 Create and assemble a composed model
31.3.10 Use the composed model

32. Document intelligence custom skill for Azure AI search
32.1 Azure AI Search enrichment pipelines
32.1.1 Indexing content in AI Search
32.1.2 What is a AI Search skillset?
32.1.3 What is a custom skill?
32.1.4 Integrate AI Search and Azure AI Document Intelligence
32.2 Building an Azure AI Document Intelligence custom skill
32.2.1 Custom skill interface and requirements
32.2.2 Testing the custom skill
32.2.3 Hosting a custom skill
32.2.4 Add the custom skill to a skillset
32.3 Lab Exercise of building and deploying an Azure AI Document Intelligence custom skill
32.3.1 Run Cloud Shell
32.3.2 Set up resources
32.3.3 Create an Azure Function
32.3.4 Configure the deployed Function
32.3.5 Publish the Function
32.3.6 Test the Function
32.3.7 Add the Function to the AI Search skillset

33. Azure OpenAI Service
33.1 Access Azure OpenAI Service
33.1.1 Create an Azure OpenAI Service resource in the Azure portal
33.1.2 Create an Azure OpenAI Service resource in Azure CLI
33.2 Use Azure AI Studio
33.2.1 Types of generative AI models
33.3 Deploy generative AI models
33.3.1 Deploy using Azure AI Studio
33.3.2 Deploy using Azure CLI
33.3.3 Deploy using the REST API
33.4 Use prompts to get completions from models
33.4.1 Prompt types
33.4.2 Completion quality
33.4.3 Making calls
33.5 Test models in Azure AI Studio's playground
33.5.1 Completions playground
33.5.2 Completions Playground parameters
33.5.3 Chat playground
33.6 Chat Completion Tuning exercise
33.6.1 Use the Chat playground
33.6.2 Experiment with system messages, prompts, and few-shot examples
33.6.3 Experiment with parameters
33.6.4 Deploy your model to a web app

34. Building Natural Language Solutions with Azure OpenAI Service
34.1 Integrate Azure OpenAI into your app
34.1.1 Create an Azure OpenAI resource
34.1.2 Choose and deploy a model
34.1.3 Authentication and specification of deployed model
34.1.4 Prompt engineering
34.1.5 Available endpoints
34.2 Use Azure OpenAI REST API
34.2.1 Chat completions
34.2.2 Embeddings
34.3 Azure OpenAI SDK
34.3.1 Install libraries
34.3.2 Configure app to access Azure OpenAI resource
34.3.3 Call Azure OpenAI resource
34.4 Lab Exercise – Integrating Azure OpenAI to your Application
34.4.1 Provision an Azure OpenAI resource
34.4.2 Deploy a model
34.4.3 Prepare to develop an app in Visual Studio Code
34.4.4 Configure your application
34.4.5 Add code to use the Azure OpenAI service
34.4.6 Test your application
34.4.7 Maintain conversation history

35. Apply Prompt Engineering with Azure OpenAI Service
35.1 Understand prompt engineering
35.1.1 Considerations for API endpoints
35.1.2 Adjusting model parameters
35.2 Write more effective prompts
35.2.1 Provide clear instructions
35.2.2 Format of instructions
35.2.3 Use section markers
35.2.4 Primary, supporting, and grounding content
35.2.5 Cues
35.3 Provide context to improve accuracy
35.3.1 Request output composition
35.3.2 System message
35.3.3 Conversation history
35.3.4 Few shot learning
35.3.5 Break down a complex task
35.3.6 Chain of thought
35.4 Lab Exercise – Utilizing Prompt Engineering in your applications
35.4.1 Explore prompt engineering techniques
35.4.2 Prepare to develop an app in Visual Studio Code
35.4.3 Configure your application
35.4.4 Add code to use the Azure OpenAI service
35.4.5 Run your application

36. Coding with Azure OpenAI
36.1 Construct code from natural language
36.1.1 AI models for code generation
36.1.2 Write functions
36.1.3 Change coding language
36.1.4 Understand unknown code
36.2 Complete code and assist the development process
36.2.1 Complete partial code
36.2.2 Write unit tests
36.2.3 Add comments and generate documentation
36.3 Fix bugs and improve your code
36.3.1 Fix bugs in your code
36.3.2 Improve performance
36.3.3 Refactor inefficient code
36.4 Lab Exercise - Generating and Improving Code with Azure OpenAI Service
36.4.1 Generate code in chat playground
36.4.2 Prepare to develop an app in Visual Studio Code
36.4.3 Configure your application
36.4.4 Add code to use your Azure OpenAI service model
36.4.5 Run the application

37. Generate Images with Azure OpenAI Service
37.1 About DALL-E
37.2 Explore DALL-E in Azure AI Studio
37.3 Lab Exercise – Generating Images with DALL-E model
37.3.1 Provision an DALL-E model
37.3.2 Explore image-generation in the images playground
37.3.3 Use the REST API to generate images
37.3.4 Prepare to develop an app in Visual Studio Code
37.3.5 Configure your application
37.3.6 View application code
37.3.7 Run the application

38. Retrieval Augmented Generation (RAG) with Azure OpenAI Service
38.1 Concepts of Retrieval Augmented Generation (RAG) with Azure OpenAI Service
38.1.1 Fine-tuning vs. RAG
38.2 Add your own data source
38.2.1 Connect your data
38.3 Chat with your model using your own data
38.3.1 Token considerations and recommended settings
38.3.2 Using the API
38.4 Lab Exercise – Implementing Retrieval Augmented Generation with Azure OpenAI Service
38.4.1 Provision Azure resources
38.4.2 Upload your data
38.4.3 Deploy AI models
38.4.4 Create an index
38.4.5 Prepare to develop an app in Visual Studio Code
38.4.6 Configure your application
38.4.7 Add code to use the Azure OpenAI service
38.4.8 Run your application